In current wireless networks, such as long term evolution (LTE) networks, network elements may be arranged in a type of hierarchy, with one upstream network element providing traffic to several downstream network elements. For example, one core network element may transmit data to several non-core or edge network elements. When the downstream network elements begin to receive too much traffic to handle and become overloaded, the performance of such elements may deteriorate and thereby adversely affect customer service. However, there is typically no means for the downstream network element to notify the upstream network element of this overload condition. The upstream network element may have access to alternate routes for providing the traffic to the end user, but since the upstream network element is not aware of the overload condition on the downstream element, if the downstream element is otherwise the best device to receive the traffic for transit to the end user, the upstream element will continue to send traffic to the overloaded downstream element.
In many wireless (and other) networks, a server-centric monitoring system where a server is statically configured to collect a limited set of data from network elements in order to determine network conditions and performance. In such configurations, the individual network elements are not capable of intelligently collecting and/or providing monitoring information to a monitoring system. The monitoring information that is in use is not granular but rather device or system level (i.e., not information regarding individual user devices and/or calls). Moreover, such systems do not allow for effective, automatic overload protection or mitigation. Some efforts in place to attempt to address these deficiencies of the current state of the art include using traces and third-party probe systems, but these efforts are not easily implemented for every network element in a network and are too costly and invasive to run continuously on an around-the-clock basis.